Displaying 1 result from an estimated 1 matches for "pn_a".
Did you mean:
p_a
2008 Mar 06
0
Help with colinearity problem in multiple linear regression
...P2_X
P2_b = t(P2_X) %*% P2[,"Murder"]
# repeat for partitions 3:N
# ...
# collect data from the all the partitions, this is the key for
# the parallelism technique, I don't know how to do this step if
# I was somehow doing QR decomposition.
A = P1_A + P2_A # ... + P3_A ... + PN_A
b = P1_b + P2_b # ... + P3_b ... + PN_b
# calculate regression, this fails because of sigularity
solve(A) %*% b
# If I exclude the introduced column it works, but I'm not
# sure how this would be generalized.
solve(A[-5,-5]) %*% b[-5]
# Compare to lm()
lm(Murder~UrbanPop+Assault+Rape+I...